Alnitak34 🛡️⚙️

Put a bounty behind your agent’s rule. Give others a reason to find the case your tests missed. I built Dissent for @monad Metropolis: a protocol that rewards valid counterexamples to measurable agent rules, with deterministic verification and onchain settlement.

I’ve built smart contracts on Monad. This update gives me another reason to look under the hood. Category Labs is adapting Monad’s execution client to produce cryptographic proofs of Ethereum blocks. Others can verify the result without replaying every transaction. As a builder, I find that connection interesting: the technology underneath the network I’m building on being adapted for use beyond it. Still in development. Worth understanding.

Beyond poker: an agent payment rule. In a coordinated Mainnet demo with Kanmani, an escrow payment was deliberately released before delivery was recorded onchain. Dissent accepted the challenge. Withdrawal: 0.75 MON bounty + 0.10 deposit returned. monadvision.com/tx/0x76525e5…

Dissent in action on Monad Mainnet. Case D-001: a challenger’s evidence returned 1 against a rule limit of 0. The contract settled the bounty. The challenger withdrew 3.10 MON: 3.00 reward + 0.10 deposit returned. Withdrawal: monadvision.com/tx/0xf2378b1…

I built Dissent for Metropolis. Live on Monad Mainnet. Agent builders can put a bounty behind a numerical rule. Find a valid counterexample, prove it onchain, and the contract pays the reward. Explore the case: dissent-henna.vercel.app/

The contracts, integration guide and recorded Mainnet case are public. Explore the evidence, inspect the code and assess whether one of your agent’s rules fits. Case: dissent-henna.vercel.app/ Code: github.com/alnitak34/dissent

Treasury spending limits. Execution thresholds. Infrastructure policies. These are possible future integrations. Each needs its own deterministic recomputer. The contribution is a reusable challenge mechanism that different agent builders can integrate.

The reference case uses Alnitak, my poker agent. A challenger presented a permitted hand that broke its registered rule under the recomputer’s model. The contract credited the bounty, and the challenger withdrew it on Monad Mainnet. Every step has a receipt.

A team can use Dissent to invite outside scrutiny with a funded reward. Define the rule. Specify the permitted evidence. Register the recomputer that checks it. Dissent provides the contracts for commitments, challenges and payouts.

What if the next AI agent you build does more than answer prompts? @devfun launched Agent TCG with a $20,000 prize pool. You bring a 60-card deck and the code that plays it. Every match puts the agent under pressure. It has to choose, adapt, risk credits and prove its strategy against other agents. TCG Stakes Season 3 is now live. The leaderboard is moving. The arena is open. Build the agent. Enter the arena.